1) I made a general case that applies to titles that are badly optimized and rushed. Picking what appears as one of the best optimized title of the year is just about cherry picking.
I do not think this is cherry picking, but one of many titles that crushes the base PS4 in performance / IQ settings on a 970 based machine. The fact that it also best PS4pro is just a nice cherry on the top for the comparisons with the pro.
1bis) Overclocking can't get into this argument as it's variable and with its own risks and downsides. We also don't know how well the game would run on an uncapped framerate on the PS4 Pro, it might as well run to an average of 40-50. We don't know. And we also don't know if this limit is more due to the CPU rather than the GPU.
It is the ability of PC hardware and something Maxwell GPUs do extremely well (I am +225 on my Maxwell core and +550 on the memory). Running for more than 2 years now.
2) What I complained about here is that DF is making comparisons without firstly establishing what graphic settings the PS4/PS4 Pro runs at. It's all about guesswork. You don't make a "technical analysis" based on guesswork. Maybe the conclusion is the same, but the analysis is poor.
They will probably have the better way to make the comparison of exact graphical settings when the game actually launches and they can run them side by side or use the dual input option they have. I honestly trust Richard though to not make the comparison to PS4Pro in bad faith though in this current video.
Instead about the 970/PS4/PS4 Pro comparison I can't remember right now a graphically intensive game that came out this year. Horizon is an exclusive. What else came out at this level? Nier Automata as far as I remember had lots of problems, running at worse LODs than even on PS4.
Tons of games have come out this year that look great and are AAA: Prey, The Surge, Tom Clancy's Ghost Recon Wild Lands, For Honor, etc. They all run really well. You mentioning Nier AUtomata for some reason, a
slightly botched port which is generally not known for its technical graphical prowess on
any platform, is quite the transparent argument.
The 970 is definitely holding up better than I expected, but I still believe the gap is narrowing and I'd like to see a comparison of Battlefront or COD or Farcry 5 to see how they run on a 970 at *console settings* (and not just "guesswork").
You expected the 970 to be equalised and bested by base PS4, which is absurd, own up to it IMO. In terms of comparisons with PS4 and PS4Pro at console graphical settings, this channel may interest you greatly. It shows the 970 performing basically like you would expect it to, providing PS4 graphics at 60 fps (with some extra bells and whistles) or PS4Pro like performance, or even better than PS4pro performance.
Analogue Foundry
What about last year Battlefield 1 and COD? I do remember COD ran very poorly on PC.
I have heard nothing of the sort about it running poorly. We know that the console version of the game has very specific "optimisations" the PC version does not make available. The consoel versions have a dynamic resolution based upn MSAA and dynaic/artist tagged alpha transparency resolutions. So getting a PC running at "console settings" is not exactly do able, since it is shifting the internal resolution of various internal buffers below native resolution constantly on consoles.
No this wasn't true. It was a problem when combining certain graphical settings that the ps4 didn't have. It was a problem with the MSAA & AO if I recall correctly.
Yep.